Merge tag 'nfs-rdma-for-5.1-1' of git://git.linux-nfs.org/projects/anna/linux-nfs

NFSoRDMA client updates for 5.1

New features:
- Convert rpc auth layer to use xdr_streams
- Config option to disable insecure enctypes
- Reduce size of RPC receive buffers

Bugfixes and cleanups:
- Fix sparse warnings
- Check inline size before providing a write chunk
- Reduce the receive doorbell rate
- Various tracepoint improvements

[Trond: Fix up merge conflicts]
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>

7 files changed, 35 insertions, 9 deletions

diff --git a/net/sunrpc/xprtrdma/backchannel.c b/net/sunrpc/xprtrdma/backchannel.c
index 749452724e6e..d79b18c1f4cd 100644
--- a/net/sunrpc/xprtrdma/backchannel.c
+++ b/net/sunrpc/xprtrdma/backchannel.c
@@ -123,7 +123,7 @@ static int rpcrdma_bc_marshal_reply(struct rpc_rqst *rqst)
 
 	rpcrdma_set_xdrlen(&req->rl_hdrbuf, 0);
 	xdr_init_encode(&req->rl_stream, &req->rl_hdrbuf,
-			req->rl_rdmabuf->rg_base);
+			req->rl_rdmabuf->rg_base, rqst);
 
 	p = xdr_reserve_space(&req->rl_stream, 28);
 	if (unlikely(!p))
diff --git a/net/sunrpc/xprtrdma/frwr_ops.c b/net/sunrpc/xprtrdma/frwr_ops.c
index 6a561056b538..52cb6c1b0c2b 100644
--- a/net/sunrpc/xprtrdma/frwr_ops.c
+++ b/net/sunrpc/xprtrdma/frwr_ops.c
@@ -391,7 +391,7 @@ frwr_wc_localinv_wake(struct ib_cq *cq, struct ib_wc *wc)
  */
 struct rpcrdma_mr_seg *frwr_map(struct rpcrdma_xprt *r_xprt,
 				struct rpcrdma_mr_seg *seg,
-				int nsegs, bool writing, u32 xid,
+				int nsegs, bool writing, __be32 xid,
 				struct rpcrdma_mr **out)
 {
 	struct rpcrdma_ia *ia = &r_xprt->rx_ia;
@@ -446,7 +446,7 @@ struct rpcrdma_mr_seg *frwr_map(struct rpcrdma_xprt *r_xprt,
 		goto out_mapmr_err;
 
 	ibmr->iova &= 0x00000000ffffffff;
-	ibmr->iova |= ((u64)cpu_to_be32(xid)) << 32;
+	ibmr->iova |= ((u64)be32_to_cpu(xid)) << 32;
 	key = (u8)(ibmr->rkey & 0x000000FF);
 	ib_update_fast_reg_key(ibmr, ++key);
 
diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c
index d18614e02b4e..6c1fb270f127 100644
--- a/net/sunrpc/xprtrdma/rpc_rdma.c
+++ b/net/sunrpc/xprtrdma/rpc_rdma.c
@@ -164,6 +164,21 @@ static bool rpcrdma_results_inline(struct rpcrdma_xprt *r_xprt,
 	return rqst->rq_rcv_buf.buflen <= ia->ri_max_inline_read;
 }
 
+/* The client is required to provide a Reply chunk if the maximum
+ * size of the non-payload part of the RPC Reply is larger than
+ * the inline threshold.
+ */
+static bool
+rpcrdma_nonpayload_inline(const struct rpcrdma_xprt *r_xprt,
+			  const struct rpc_rqst *rqst)
+{
+	const struct xdr_buf *buf = &rqst->rq_rcv_buf;
+	const struct rpcrdma_ia *ia = &r_xprt->rx_ia;
+
+	return buf->head[0].iov_len + buf->tail[0].iov_len <
+		ia->ri_max_inline_read;
+}
+
 /* Split @vec on page boundaries into SGEs. FMR registers pages, not
  * a byte range. Other modes coalesce these SGEs into a single MR
  * when they can.
@@ -733,7 +748,7 @@ rpcrdma_marshal_req(struct rpcrdma_xprt *r_xprt, struct rpc_rqst *rqst)
 
 	rpcrdma_set_xdrlen(&req->rl_hdrbuf, 0);
 	xdr_init_encode(xdr, &req->rl_hdrbuf,
-			req->rl_rdmabuf->rg_base);
+			req->rl_rdmabuf->rg_base, rqst);
 
 	/* Fixed header fields */
 	ret = -EMSGSIZE;
@@ -762,7 +777,8 @@ rpcrdma_marshal_req(struct rpcrdma_xprt *r_xprt, struct rpc_rqst *rqst)
 	 */
 	if (rpcrdma_results_inline(r_xprt, rqst))
 		wtype = rpcrdma_noch;
-	else if (ddp_allowed && rqst->rq_rcv_buf.flags & XDRBUF_READ)
+	else if ((ddp_allowed && rqst->rq_rcv_buf.flags & XDRBUF_READ) &&
+		 rpcrdma_nonpayload_inline(r_xprt, rqst))
 		wtype = rpcrdma_writech;
 	else
 		wtype = rpcrdma_replych;
@@ -1313,7 +1329,7 @@ void rpcrdma_reply_handler(struct rpcrdma_rep *rep)
 
 	/* Fixed transport header fields */
 	xdr_init_decode(&rep->rr_stream, &rep->rr_hdrbuf,
-			rep->rr_hdrbuf.head[0].iov_base);
+			rep->rr_hdrbuf.head[0].iov_base, NULL);
 	p = xdr_inline_decode(&rep->rr_stream, 4 * sizeof(*p));
 	if (unlikely(!p))
 		goto out_shortreply;
diff --git a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
index b908f2ca08fd..907464c2a9f0 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
@@ -304,7 +304,6 @@ xprt_setup_rdma_bc(struct xprt_create *args)
 	xprt->idle_timeout = RPCRDMA_IDLE_DISC_TO;
 
 	xprt->prot = XPRT_TRANSPORT_BC_RDMA;
-	xprt->tsh_size = 0;
 	xprt->ops = &xprt_rdma_bc_procs;
 
 	memcpy(&xprt->addr, args->dstaddr, args->addrlen);
diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
index 3c490172d383..5d261353bd90 100644
--- a/net/sunrpc/xprtrdma/transport.c
+++ b/net/sunrpc/xprtrdma/transport.c
@@ -332,7 +332,6 @@ xprt_setup_rdma(struct xprt_create *args)
 	xprt->idle_timeout = RPCRDMA_IDLE_DISC_TO;
 
 	xprt->resvport = 0;		/* privileged port not needed */
-	xprt->tsh_size = 0;		/* RPC-RDMA handles framing */
 	xprt->ops = &xprt_rdma_procs;
 
 	/*
diff --git a/net/sunrpc/xprtrdma/verbs.c b/net/sunrpc/xprtrdma/verbs.c
index 21113bfd4eca..89a63391d4d4 100644
--- a/net/sunrpc/xprtrdma/verbs.c
+++ b/net/sunrpc/xprtrdma/verbs.c
@@ -1481,6 +1481,8 @@ rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, bool temp)
 	if (ep->rep_receive_count > needed)
 		goto out;
 	needed -= ep->rep_receive_count;
+	if (!temp)
+		needed += RPCRDMA_MAX_RECV_BATCH;
 
 	count = 0;
 	wr = NULL;
diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h
index 5a18472f2c9c..10f6593e1a6a 100644
--- a/net/sunrpc/xprtrdma/xprt_rdma.h
+++ b/net/sunrpc/xprtrdma/xprt_rdma.h
@@ -205,6 +205,16 @@ struct rpcrdma_rep {
 	struct ib_recv_wr	rr_recv_wr;
 };
 
+/* To reduce the rate at which a transport invokes ib_post_recv
+ * (and thus the hardware doorbell rate), xprtrdma posts Receive
+ * WRs in batches.
+ *
+ * Setting this to zero disables Receive post batching.
+ */
+enum {
+	RPCRDMA_MAX_RECV_BATCH = 7,
+};
+
 /* struct rpcrdma_sendctx - DMA mapped SGEs to unmap after Send completes
  */
 struct rpcrdma_req;
@@ -577,7 +587,7 @@ void frwr_release_mr(struct rpcrdma_mr *mr);
 size_t frwr_maxpages(struct rpcrdma_xprt *r_xprt);
 struct rpcrdma_mr_seg *frwr_map(struct rpcrdma_xprt *r_xprt,
 				struct rpcrdma_mr_seg *seg,
-				int nsegs, bool writing, u32 xid,
+				int nsegs, bool writing, __be32 xid,
 				struct rpcrdma_mr **mr);
 int frwr_send(struct rpcrdma_ia *ia, struct rpcrdma_req *req);
 void frwr_reminv(struct rpcrdma_rep *rep, struct list_head *mrs);